Shading loss calculation with multiple orientations using pvc import

[klarchet]

Hi,

I have performing simulations for complex terrain using a 3D model pvc import from PVCASE. There seems to be an underlying difference in how PVsyst calculates near shading losses depending on how you determine the orientations, either using the automatic identification or manually defining them. For an experiment, I increased the tolerance of the automatic identification tool so that only one orientation was recognised and also separately selected all the tables imported and created an orientation from manual selection. Both ways resulted in a single identical orientation so I would have expected that the near shading loss table/graph to be the same also, but this was not the case. In fact the resultant shading curve are very different (please see two attached images).

Could you explain what is the underlying difference in the near shading loss calculation between the two methods? If this is an error (I ran this experiment for both 7.2 and 7.1 versions), could you let me know which is the more representative.

auto identification

manual identification

[dtarin]

Interesting find.